ICP - 3
September 1991
AA or ICP - Which do you choose?
Geoff Tyler
Varian Australia Pty Ltd
Australia
The typical maximum temperature for an air/acetylene
flame is 2 300 °C while for nitrous oxide acetylene, it is
2 900 °C. Temperatures as high as 10 000 K can be
reached in an argon plasma.
Introduction
For many analysts Atomic Absorption Spectrometry
(AAS) is a well established and understood technique.
However, even though Inductively Coupled Plasma
Emission Spectrometry (ICP-ES) instrumentation has
been commercially available for over a decade, the
technique has proven to be more complex. This article
discusses the main differences between the two
techniques.
Detection limits
The comparison of detection limits in table 1 highlights
the following differences:
AAS vs ICP
Observation Height (mm)
The basic difference between the two techniques is that
one relies upon an atomic absorption process while the
other is an atomic/ionic emission spectroscopic
technique. The next essential difference is the means
by which the atomic or ionic species are generated. A
combustion flame or graphite furnace is typically used
for AA while ICP-ES uses a plasma.
Furnace AA detection limits are generally better in
all cases where the element can be atomized.
(ii)
Detection limits for Group I elements (e.g. Na, K)
are generally better by flame AAS than by ICP.
(iii) Detection limits for refractory elements (e.g. B, Ti,
V, Al) are better by ICP than by flame AAS.
(iv) Non metals such as sulfur, nitrogen, carbon, and
the halogens (e.g. I, Cl, Br) can only be determined
by ICP.
While it is possible to determine phosphorous by AAS,
its detection limit by ICP is more than three orders of
magnitude better.
Plasma Tail
Optimum detection of non metals such as S, N and
halogens by ICP-ES can only be achieved if a vacuum
monochromator, with purged transfer optics, is used.
The optics must be purged to exclude atmospheric
oxygen and eliminating its absorption.
25
20
15
(i)
Normal Analytical Zone
10
Sulfur can be measured at 180.73 nm by purging the
monochromator. To detect the primary aluminium
wavelength at 167.08 nm, the monochromator must
first be evacuated, then purged with the inert plasma
gas.
5
0
Initial Radiation Zone
Induction Zone
≈
Preheating
Figure 1:
Zone
Note that a continuous flow vapor generation accessory
can be used with either ICP-ES or AAS for improved
detection limits for As, Se, Hg, Sb, Bi and Ge.
A plasma used for emission spectrometry. The
regions refer to those seen when a Yttrium
solution is introduced.
1
Sample throughput
Ionization
In ICP-ES, the rate at which samples may be
determined depends on the type of instrument: both
simultaneous and sequential ICP spectrometers are
available. Most ICP spectrometers purchased are the
sequential type, providing maximum flexibility of choice
of element and analytical wavelength. Surveys have
shown that most analysts are interested in 6-15
elements per sample and choose to pump the sample
(which increases washout times) to improve precision
and accuracy by minimizing viscosity effects.
Simultaneous ICP spectrometers demonstrate an
advantage in analytical speed over sequential ICP
spectrometers when more than 6 elements/sample are
measured.
The ICP contains a large number of free electrons, so
ionization interferences for most applications are
virtually nonexistent. Ionization interferences can be
encountered when determining elements in matrices
that contain very high concentrations of Group I
elements (e.g. Na & K). However, these effects can be
minimized by optimizing the plasma viewing height.
Ionization interferences may also be found in AAS, e.g.
when measuring certain Group II elements in a nitrous
oxide flame. An ionization buffer such as Cs, Li or K
can be added to both samples and standards to
minimize this effect.
Spectral
If a ‘one off’ sample is presented for a few elements,
flame AAS is faster. However, with flame equilibration
time, program recall and monochromator condition
changes, the cross over point where sequential ICP
becomes faster than AAS is approximately 6 elements/
sample for routine analysis.
The optical requirements of AAS are fairly simple. The
monochromator only needs to distinguish a spectral line
emitted from the hollow cathode lamp from other
nearby lines. The lamp itself only emits a few spectral
lines. Most elements require 0.5 nm resolution with only
iron, nickel and cobalt of the common elements
requiring
0.2 nm or better.
Unattended operation
Flame AAS cannot be left completely unattended for
safety reasons. An ICP-ES instrument or graphite
furnace AA can be left to run overnight as no
combustible gases are involved, effectively increasing
the working day from 8 hours to 24 hours.
In ICP-ES, the rich spectra present in the plasma
means that there is a greater possibility of spectral
interference. Spectral resolutions of 0.010 nm or better
are required to resolve nearby interfering lines from the
atomic and ionic analytical emission signals of interest.
Linear dynamic range
Spectral interference in sequential ICP spectrometers
can, in most cases, be overcome by selecting a
different elemental wavelength with similar detection
limits. With simultaneous ICP spectrometers, the
elements and the wavelengths which may be
determined are fixed at the time of purchase, and an
alternative line may not be available. In this case, interelement correction may be used to minimize the
spectral interference.
The inductively coupled plasma is doughnut shaped
(with a ‘hollow’ centre). The sample aerosol enters the
base of the plasma via the injector tube. The ‘optical
thinness’ of the ICP results in little self absorption and is
the main reason for the large linear dynamic range of
about 105. For example, copper can be measured at the
324.75 nm wavelength from its detection limit of about
0.002 ppm to over 200 ppm. In ICP, extrapolation of
two point calibrations can be accurately used to achieve
orders of magnitude above the top standard. This
compares to a linear dynamic range of typically 103 for
AAS.
Physical
These interferences relate to the different properties of
various samples and can affect sample transport and
droplet formation. ICP tends to be more susceptible to
such interference because of the smaller droplet size
required and lower transport efficiency.
Interferences
Chemical
Precision
Chemical interferences are relatively common in AA,
especially with graphite furnace AA, but may be
minimized with chemical modifiers.
Precision can be termed short term (or within-run) and
long term (over a period of one day). For AAS a
precision of 0.1-1% is typical for the short term, but
recalibration is required over a longer period. With
ICP-ES the short term precision is typically 0.3-2%, but
precisions of 2-5% are not uncommon over an 8 hour
period without recalibration.
ICP-ES is almost free from chemical interferences. The
chemical bonds that still exist at below 3000 °C are
completely ruptured at above 6000 °C. The high
temperatures reached in a plasma eliminate chemical
interferences, which accounts (for the most part) for the
better detection limits achieved for refractory elements.
2
One technique used to eliminate backlash in the grating
drive mechanism of ICP spectrometers is by scanning
and measuring at the same time. This method of
measurement can be termed as ‘measurement on the
move’ and effectively results in poor short term
precision. A more recent method drives the grating to a
wavelength near the analytical peak. A refractor scan is
then performed over a smaller wavelength region in
order to identify and locate the peak position. Finally the
refractor plate is repositioned ‘at the peak’ where the
replicate measurements are then performed. This
method offers better precision.
(d) What are the typical sample volumes?
(e) What elements need to be determined?
(f)
What concentration ranges are present in the
matrices?
(g) Would an Internal Standard be useful? For
example, where the samples may change in
viscosity from sample to sample, e.g. battery acid
analysis.
(h) What expertise do the operators have?
(i)
How much money is available to purchase or lease
costs/month?
(j)
Cost of ownership and running costs. Can the user
afford an automated AAS or ICP-ES, or is a simple
AAS sufficient?
Analytical requirements
Before deciding which technique is appropriate, the
chemist must define both present and future analytical
requirements. That is:
The answers to these questions will help you to decide
which is the preferred technique. Sometimes the answer
is further complicated by the fact that neither flame AAS
nor ICP-ES will satisfy all requirements. You may find,
as many do, that both an ICP-ES and a furnace AAS
will be necessary to meet the analytical requirements.
(a) Number of samples/week?
(b) What matrices need to be analyzed? e.g. steels,
bronzes, effluents, soils, etc.
(c) How many elements need to be determined for
each sample type?
3
AAS v ICP - A quick guide
ICP-OES
Detection
Flame AAS Furnace AAS
Best for :
Limits Refractories
Best for :
Best for :
Group I metals
All elements
Non metals
Na, K
except : B,W,U,
P, S, B, Al
Volatile elements
Refractories, eg
V, Ba, Ti
Pb, Zn
P, S Halogens
Rare Earths
Sample
Best if more than
Best if less than
Slow (typically
Throughput
6 elements/sample
6 elements/sample
4 mins/element)
Linear Dynamic
Range
105
103
102
0.3 - 2%
Less than 5%
0.1 - 1%
0.5 - 5%
Interferences
Spectral
Many
Virtually None
Minimal
Chemical
Ionization
Virtually None
Minimal
Some
Some
Many
Minimal
Low
Yes
Relatively high
No
Precision
Short Term
Long Term
(over 8 hrs)
Operating costs
High
Combustible gases No
4
Table 1
Guide to ICP/AAS Analytical Values
Element
ICP
λ (nm)
AA
Silver
Ag
Aluminium
Al
Arsenic
As
Gold
Au
Boron
B
Barium
Ba
Beryllium
Be
Bismuth
Bi
Bromine
Br
Carbon
C
Calcium
Ca
Cadmium
Cd
Cerium
Ce
Chlorine
Cl
Cobalt
Co
Chromium
Cr
Cesium
Cs
Copper
Cu
Dysprosium
Dy
Erbium
Er
Europium
Eu
Iron
Fe
Gallium
Ga
Gadolinium
Gd
Germanium
Ge
Hafnium
Hf
Mercury
Hg
Holmium
Ho
Iodine
I
Indium
In
Iridium
Ir
Potassium
K
Lanthanum
La
Lithium
Li
Lutetium
Lu
Magnesium
Mg
Manganese
Mn
Molybdenum Mo
Nitrogen
N
Sodium
Na
Niobium
Nb
Neodymium
Nd
Nickel
Ni
Osmium
Os
Phosphorous P
Lead
Pb
Palladium
Pd
Praseodymium Pr
Platinum
Pt
Rubidium
Rb
Rhenium
Re
328.1
309.3
193.7
242.8
249.8
553.6
234.9
223.1
422.7
228.8
520.0
240.7
357.9
852.1
324.7
421.2
400.8
459.4
248.3
294.4
368.4
265.1
307.3
253.7
410.4
303.9
208.9
766.5
550.1
670.8
336.0
285.2
279.5
313.3
589.0
334.9
492.5
232.0
290.9
213.6
217.0
244.8
495.1
265.9
780.0
346.1
ICP
328.068
167.081
188.985
267.595
249.773
455.403
313.042
223.061
163.340
247.856
393.366
228.802
418.660
725.665
228.616
267.716
455.531
324.754
353.170
337.271
381.967
259.940
417.206
342.247
265.118
264.141
184.950
345.600
178.276
325.609
224.268
766.490
379.478
670.784
261.542
279.553
257.610
202.030
174.272
588.995
309.418
401.225
231.604
225.585
177.499
220.353
340.458
417.939
265.945
780.023
227.525
Flame AA
Zeeman Furnace AA
Detection
Characteristic
Detection
Limit
Conc
Limit
µg/L
µg/L
µg/L
3
1.5
12
5.5
1.5
0.07
0.2
12
6000
65
0.03
1.5
7.5
200000
5
4
3200
2
0.3
0.7
0.3
1.5
6.5
2.5
13
4
8.5
0.5
60
18
3.5
10
0.02
0.6
0.05
0.1
0.3
4
50 000
1
4
2
5.5
5
18
14
7
0.8
20
35
11
FlameCharacteristic#
Type
MSR
Conc+
Mass
µg/L
pg
%
El
97
100
86
94
70
100
64
88
Ag
Al
As
Au
B
Ba
Be
Bi
Br
C
Ca
Cd
Ce
Cl
Co
Cr
Cs
Cu
Dy
Er
Eu
Fe
Ga
Gd
Ge
Hf
Hg
Ho
I
In
Ir
K
La
Li
Lu
Mg
Mn
Mo
N
Na
Nb
Nd
Ni
Os
P
Pb
Pd
Pr
Pt
Rb
Re
30
800
500
100
8000
200
15
200
2
30
300
10
500
20
1
50
Air
N2O
N2O
Air
N2O
N2O
N2O
Air
0.035
0.25
0.5
0.22
43
0.85
0.025
0.45
0.7
5
10*
4.4
855*
17
0.5
9
10
10
100000
1
2
100000
N2O
Air
N2O
0.03
0.01
0.6
0.2*
50
50
20
30
600
500
300
50
800
20000
1000
10000
1500
700
5
6
4
3
30
50
1.5
6
100
2000
200
2000
200
40
Air
N2O
Air
Air
N2O
N2O
N2O
Air
Air
N2O
N2O
N2O
Air
N2O
0.21
0.075
0.55
0.3
2.3
5
1.3
0.06
0.23
4.2
1.5
11
6
45
100
25
1.2
4.5*
0.45
9*
7.5
150*
150
800
7
40000
20
7000
3
20
300
40
500
3
2000
2
300
0.3
2
20
Air
Air
Air
N2O
Air
N2O
Air
Air
N2O
0.35
6.8
0.02
7.0*
135
0.4
0.2
4
0.01
0.03
0.35
0.2
0.6
7
3
20000
6000
70
1000
120000
100
50
20000
1000
50
8000
0.2
2000
1000
10
100
40000
10
10
10000
100
10
1000
Air
N2O
N2O
Air
N2O
N2O
Air
Air
N2O
Air
Air
N2O
0.005
0.1
5
0.24
110
0.28
0.43
3.5
0.05
94
87
98
100
58
84
100
100
100
97
80
100
69
100
97
90
49
75
92
96
92
4.8
98
2200* 69
5.5
92
8.6
100
70
82
1
90
-
Guide to ICP/AAS Analytical Values
Element
ICP
λ (nm)
AA
Rhodium
Ruthenium
Sulphur
Antimony
Scandium
Selenium
Silicon
Samarium
Tin
Strontium
Tantalum
Terbium
Tellurium
Thorium
Titanium
Thallium
Thulium
Uranium
Vanadium
Tungsten
Yttrium
Ytterbium
Zinc
Zirconium
Rh
Ru
S
Sb
Sc
Se
Si
Sm
Sn
Sr
Ta
Tb
Te
Th
Ti
Tl
Tm
U
V
W
Y
Yb
Zn
Zr
343.5
349.9
217.6
391.2
196.0
251.6
429.7
235.5
460.7
271.5
432.7
214.3
364.3
276.8
371.8
358.5
318.5
255.1
410.2
398.8
213.9
360.1
ICP
343.489
267.876
180.734
217.581
361.384
196.026
251.611
442.434
242.949
407.771
268.517
350.917
214.281
274.716
334.941
351.924
346.220
385.958
309.311
239.709
371.030
328.937
213.856
339.198
Flame AA
Zeeman Furnace AA
Detection
Characteristic
Detection
Limit
Conc
Limit
µg/L
µg/L
µg/L
5
5.5
20
18
0.4
37
5
7
15
0.02
9
5
27
17
0.6
16
1.5
18
2
17
0.2
0.3
0.9
1.5
FlameCharacteristic#
Type
MSR
Conc+
Mass
µg/L
pg
%
El
95
100
96
92
100
93
94
90
93
100
63
79
97
92
-
Rh
Ru
S
Sb
Sc
Se
Si
Sm
Sn
Sr
Ta
Tb
Te
Th
Ti
Tl
Tm
U
V
W
Y
Yb
Zn
Zr
100
400
5
100
Air
Air
0.4
0.75
8
15
300
300
1000
1500
6000
700
40
10000
7000
200
40
50
500
300
1000
100
2
2000
700
30
Air
N2O
N2O
N2O
N2O
N2O
N2O
N2O
N2O
Air
0.5
10
0.7
0.75
14*
15
0.5
0.1
10*
2
0.18
0.45
3.5
9*
1000
200
300
100000
700
5000
2000
60
8
9000
100
20
20
40000
100
1000
200
4
1.0
1000
N2O
Air
N2O
N2O
N2O
N2O
N2O
N2O
Air
N2O
2.5
0.75
50
15
1.1
22
0.15
3
0.0075 0.15
NOTES:
* Modifier used to obtain these results.
+ 20 µL injection
# The Characteristic Masses listed were determined in aqueous solution using maximum heating rate in argon
with zero gas flow during atomization.
For Deuterium Furnace systems, the equivalent Characteristic Concentration and Characteristic Mass is
easily calculated using the following conversion:
CMn = CMz X MSR (%)/100
CCn = CCz X MSR (%)/100
where:
CMn = Characteristic Mass for Deuterium Furnace Systems
CMz = Characteristic Mass for Zeeman Furnace Systems (from Table above)
MSR = Magnetic Sensitivity Ratio (as % from Table above)
CCn = Characteristic Concentration for Deuterium Furnace Systems
CCz = Characteristic Concentration for Zeeman Furnace Systems (from Table above).
6